Figure 71. Padden Creek pocket estuary in 1888 (previous page), and in 2004 (above).
Estuarine Fauna Estuaries provide a range of habitats that are exceedingly complex. These include edge, bottom and open water environments. Significant environmental factors to salmon include water depth, salinity, temperature, turbidity, and velocity that vary according to seasonal, lunar, and tidal time scales (Quinn 2005). These environmental factors affect the where and when organisms reside in the estuary. For example, early in the salmon out-migration season, water temperatures in the Nooksack River are cooler than the adjacent marine waters. During the mid-season period, these temperatures become similar. Towards the late season, on the outer delta, the upper layers of stratified marine
123
water become warmer than the river water, at times approaching the upper lethal limit for salmon. Tidal influences affect high water temperatures, which are more likely during minus tides following the lunar (28 day) cycle. On a diurnal scale, high water temperatures are greatest following a minus tide that occurs during the mid-day because of the transfer of heat energy from the tide flats to the water as the tide comes in. Under these conditions, juvenile salmon may choose to remain in the upper portion of distributary channels influenced by cooler river water, or move into deeper marine waters.
Estuaries are less numerous and more different from each other than other types of aquatic habitat such as those found in stream environments. As a result, they are less amenable to comparative research approaches (Quinn 2005). Empirical data gathered on fish habitat utilization is perhaps more relevant in developing restoration strategies for estuaries than for habitat found in other environments.
Organisms found in the Nooksack River estuary, whether temporary or permanent residents, need energy in the appropriate forms to survive. Plants require light and nutrients, and serve as primary producers. Herbivorous animals (salmon prey) are often referred to as primary-level consumers. Carnivorous animals that eat herbivores are called secondary-level consumers; carnivores preying on these secondary consumers are called tertiary, or higher level consumers (Thom 1987). Pacific salmon species display feeding habits that utilize both secondary and tertiary level consumers, depending on their life stage. Chinook salmon in the estuary as early-age fish feed on herbivorous invertebrates. Later, as larger fry and fingerlings, they supplement their diet by feeding on larval and juvenile fish species, as well as insects (Thom 1987, Simenstad et al.1982). The fork length (FL) size of juvenile salmon determines to a large extent the choice of available prey. Larger individuals are faster swimmers capable of capturing larger range of prey species. As a result, the size of salmon prey shares a positive linear relationship with a salmon’s body size.
The food web in the estuary is highly complex (Figure 72). Processes that maintain the food web are similar to those that differentiate the vegetation zones in the estuary. Salinity gradients, a direct result of hydrological processes in the estuary, influence the abundance, diversity and distribution of primary and secondary consumers, just as they do vegetation assemblages. Sediment accumulation and distribution, the result of hydrology, wind, and wave energy also influence the distribution of estuarine food web items.
124
Figure 71. The estuarine food web (WADOE 2004).
Hydrology in the estuary affects the abundance and break down of food resources. Freshwater entering the estuary deliver and remove nutrients and food items (Allan 1996). The current velocity in estuarine channels has an effect on the distribution of substrate particle sizes. This in turn influences invertebrate size and distribution. Estuarine marsh invertebrate assemblages are regulated by or coincident with marsh soil development and detritus trapping (Simenstad and Cordell 2000). The flow regime that
125
alters the construction and composition of the substrate affects the buildup of detritus. A lack of detritus reduces levels of salmon prey detritivores. This inhibits higher trophic levels from establishing communities. Juvenile salmon prey such as harpacticoid copepods feed on bacterial flora associated with organic detritus. Detritus also creates conditions that retain moisture, providing refuge to invertebrates during periods that are dewatered by retreating tides.
Macroinvertebrates Macroinvertebrates are organisms that are large (macro) enough to be seen with the naked eye and lack a backbone (invertebrate). They inhabit all types of environments, from fast-flowing mountain streams and slow-moving muddy rivers, to terrestrial canopies. Benthic macroinvertebrates are organisms that live on the channel bottom.
Benthic macroinvertebrate communities are integral components of both freshwater and estuarine systems. These organisms live on or within sediments; influence sediment and water chemistry; alter sediment organic content and structure; and serve as major prey species for many species of fish, including some salmon species (Cuomo and Zinn 1997). Valuable low-level residents in the estuarine food web (Figure 72), they constitute a large part of the diet of juvenile salmonids (Schabetsberger et al. 2003, Cuomo and Zinn 1997, Hayman et al. 1996, Hart 1980). Estuarine benthos typically includes nematode worms, polychaete worms, amphipods, isopods, copepods, gastropods, and marine mollusks (Cuomo and Zinn 1997).
Rhoads and Boyer (1982, cited in Cuomo and Zinn 1997) documented a series of predictable stages for the development of benthic communities. These successional sequences (Stage I, II and III) are characterized by particular, functional types of benthic organisms. One functional type succeeds another over time if all else remains stable. Organisms comprising Stage I estuarine assemblages colonize newly available seafloor, like freshwater habitat recently inundated by saline water. A change in benthic community composition from freshwater benthos to a Stage I estuarine benthic assemblage can be expected to occur with salt marsh restoration. As saline estuarine waters are introduced into a freshwater marsh environment, freshwater organisms such as chironomids, leeches, and oligochaetes will be replaced by salt-tolerant Stage I organisms such as polychaete worms and amphipods.
If no disturbance occurs to reset the successional process, the intermediate Stage I community will eventually be succeeded by organisms such as polychaete worms, bivalve shellfish, and organisms that burrow deeper into sediments (Cuomo and Zinn 1997). Estuarine sediments in higher energy systems usually support only Stage I and early Stage II communities because they are often subjected to frequent sediment disturbance (McCall 1978, cited in Cuomo and Zinn 1997).
Simenstad and Cordell (2000) list macroinvertebrate composition and density in the estuary as one of the primary success criteria employed when assessing estuarine habitat function. The measurement of invertebrate diversity and distribution is an important data set that offers insight into what the Nooksack estuary offers juvenile migrating juvenile
126 salmon migrating during their critical transformation from freshwater to marine individuals.
In 2003 and 2004 LNR staff and student interns from Northwest Indian College and Western Washington University’s Huxley College conducted macroinvertebrate prey sampling during the peak Nooksack basin chinook estuarine migratory period (January through June). The assessment objective was to collect baseline data on the abundance and distribution of benthic macroinvertebrate species within all habitat zones of the Nooksack River’s lower watershed and estuary as bioindicators of fish habitat health. Although benthic invertebrates are only one of several food resources utilized by juvenile salmonids in the estuary, they are easily contained, stable in their habitat selection and do not migrate throughout the estuary, and are excellent indicators of water quality and habitat health. This project will serve as a template for further monitoring of benthic communities. Benthic sampling for macroinvertebrate food resources will assist future salmon restoration efforts by identifying areas with high prey species diversity and productivity. This information will assist project development by emphasizing and improving juvenile salmonid utilization of habitats with notable food resources.
Benthic macroinvertebrates were sampled during their winter and spring biological phases at 24 sites located within the Nooksack and Lummi River watersheds and their associated nearshore habitats (Figure 73). Sampling site locations were chosen to be a representative sample of all microhabitat zones available to juvenile salmonids during estuarine residence. Sample sites represented the many different habitats available to salmonids in the Nooksack estuary, from freshwater tidal mainstem, to brackish salt marsh, to highly saline nearshore. Phase-1 sampling occurred between January and February, Phase-2 sampling occurred between March and April, and Phase-3 occurred between May and June. The majority of the samples were collected from a boat, while five samples were collected from bridges, and two from shore. Sediment composition greatly influences invertebrate composition; therefore, sediments also collected in the sample were characterized.
127
Figure 72. Macroinvertebrate sampling sites in the Nooksack River estuary, 2003 and 2004. Site labels correspond to geographic area: L – Lummi River and delta, P – Portage Island and Bay, and N – Nooksack River and delta; and to sites described in Table 6.
Benthic samples were collected with a 6 x 6 x 6 inch Eckman dredge (Figure 74). Total sample volume varied with substrate composition. The average sample volume was approximately 105 in3 as a result of variation in bottom hardness (Ross and Weispfenning 2004). Gravel and coarse sand substrates typically had the lowest sample volume, while silt and mud substrates had the greatest sample volume. One dredge load was removed from each site during each sampling phases. Invertebrates, sediments, macrophytes, and debris collected in each sample were preserved in the field, and sorted and characterized in the lab. Each sample was thoroughly washed on a 1 mm sieve to remove fine sediments and qualitative notes were made on the composition of the substrate. The sample was carefully transferred to a 500 mL Nalgene container and fixed with 10 % buffered formalin solution for 72 hours. The formalin was then transferred to a toxic waste container and the sample preserved in 70 % ethanol.
128
Figure 73. Lummi Natural Resources field technician using the Eckman dredge to sample benthic macroinvertebrates in the Lummi River, 2003.
Mollusks and casings were separated out so soft-bodied organisms consumed by fish would not be overstated. Dry weight biomass of soft-bodied invertebrates, which excluded mollusks and the tubes of polychaetes, was determined by desiccating samples at 60 °C for 24 hours. Species diversity by site was calculated using the Shannon-Weiner Index and the Pielou’s Evenness Index, and overall abundance by site was summed from all samples collected (Ross and Weispfenning 2004).
The variability of invertebrate distribution within the water column is an important consideration when analyzing sampling data. Diurnal periodicity may influence the migration of mobile species. All benthic sampling was conducted during daylight hours, for boat safety reasons. This sampling schedule may influence our findings of species abundance. Another important effect on distribution of invertebrates is cover. Several species of benthic dwellers that were collected at the benthic surface in the estuary were also found mobile within wood and rock assemblages; these hard, inconsistent surfaces could not be sampled with the dredge. Traps with uniform substrate surfaces reduce this variability, but were deployed later in the sampling period after it was realized that several sampling sites could not adequately be sampled during high flows with the sampling gear. Results from this alternative sampling method are pending, and not included as data in this section.
The following groups comprise the majority of benthic invertebrate food items consumed by Puget Sound juvenile salmon during their estuarine and nearshore residencies (Brennan 2004, Simenstad et al. 2003, Schabetsberger et al. 2003, Levy et al. 1979, Bailey et al. 1975, Dunford 1975). A complete list of invertebrate species collected in the
129
Nooksack estuary and nearshore environment is available in Appendix A tables 10, 12, 13, and 14.
Phylum: Arthropoda; Class: Crustacea; Order: Copepoda Copepods are small crustaceans that lack compound eyes and a carapace. Most copepods are between 0.5 – 2.0 mm long, serving as an important food source for young salmonids, forage fish and invertebrates.
Benthic and epibenthic harpacticoid copepods (Harpacticus sp.) are important prey items for juvenile salmonids, and were found primarily in nearshore habitats on either side of the Nooksack delta tide flat. Several harpacticoids live out of water, on salt marsh habitat, but enter aquatic habitats if an incoming tide pulls them off of land in to the water. The depth at which these copepods swim depends not only on the species and sex, but also on the temperature of the water, the season, the hour of the day and amount of light present (Kaestner 1980). In general, copepods rise toward the surface in the late afternoon as a result of swimming toward the light source of decreasing intensity. This upward migration is continued into the night, oriented by gravity; vertical migrations range from a few meters to 150 m or more (Kaestner 1980).
Phylum: Arthropoda; Class: Crustacea; Order: Amphipoda Amphipods live in a variety of estuarine environments, from low-flow tidal channels and salt marsh flats, to rocky and sandy intertidal communities. Juvenile chinook and coho salmon diets sampled by Schabetsberger (et al. 2003) in the Columbia River estuary plume regularly consisted of hyperiid amphipods, along with larval fish, crab megalopae, and euphausiids (crustacean krill). Forage fish species important to juvenile salmon in the estuary also feed heavily on amphipods (WDOE 2004).
Corophium sp. is an amphipod that contributes significantly to the diet of migrating juvenile salmon (Salamunovich 1987). They comprised the majority of invertebrates found in brackish environments in the winter and early spring in the Nooksack estuary, and in more freshwater environments later in the salmonid migration period.
Phylum: Arthropoda; Class: Crustacea; Order: Isopoda Isopods are small crustaceans that have flat bodies and eight legs. They primarily eat detritus and marine vegetation. They are most commonly found in low energy nearshore environments and eelgrass beds. Pennings (et al. 2000) found that one estuarine isopod, Ligia pallasii, tended to prefer wrack (aged, stranded seaweeds) to fresh seaweeds of the same species in Pacific Northwest studies. These results suggest that increased organic and mineral contents of marine drift and the eventual build-up of detritus is important in the diets of primary feeders in estuaries. Marine isopods in all life stages are consumed by juvenile coho and chinook, and Dungeness crab in nearshore habitats.
Phylum: Annelida; Class: Polychaeta, Oligochaeta Annelids are segmented worms that crawl over or burrow into soft sediment surfaces. The vast majority of the more than 8,000 known species of polychaete worms are marine; some, however, are found in fresh or brackish water. They are abundant from the intertidal zone to depths of over 16,405 ft (5,000 m). The polychaetes, so named because
130
of the numerous setae they bear, range in length from less than 1/8 in. to more than 9 ft (2 mm to 3 m), but most are from 2 to 4 in. (5–10 cm) long (Columbia Encyclopedia 2004).
There are about 3,500 species of oligochaete earthworms and freshwater worms. The members of this class range in length from about 1/32 in. to 10 ft (0.5 mm – 3 m), but most are comparable to the polychaetes in size. Oligochaetes occur in a variety of habitats throughout the world. Most are burrowers in the soil, but the class also includes worms that inhabit wells, marshes, and swamps. Other species live under rocks on the seashore, in the leaves of trees and vines, or on the gills of freshwater crayfish (Columbia Encyclopedia 2004).
Annelid worms are a keystone species in the estuary and nearshore. They serve as part of the detrital food web. They are easily digestible, and consumed by all juvenile salmon and trout species.
Phylum: Mollusca; Class: Pelecypoda (Bivalvia), Gastropoda Gastropods, the largest class of mollusks, include limpets, top shells, snails, slugs, sea hares, abalones, and nudibranchs, or sea slugs. They are found mainly in brackish and saline environments; several species take advantage of the lower energy environments of the estuary and nearshore for spawning and rearing. Gastorpods in the larval stage are most easily assimilated by young salmon (Columbia Encyclopedia 2004).
Species Biomass The biomass from collected samples represents the relative potential at each site for nutrient contributions to salmon and other higher food web organisms. Biomass is expressed in Total Dry Weight in grams. The Total Dry Weight of macroinvertebrates per site was uniformly highest across all sampling sites during March-April Phase 2 (1.955 g). Biomass from Phase 3 (1.66 g), and phase 1 (1.543 g) samples decreases in magnitude. Figure 75 summarizes the Total Dry Weight at each sampling site during each sampling phase. Of the 24 sampling sites, L7 (Sandy Point Nearshore) had the highest total combined biomass from all 3 sampling phases (1.456 g) (Table 6). The Lummi River and Portage Island sites tended to have consistently high total biomass over all three sampling phases. The Nooksack River sites had very low biomass in comparison to the Lummi River and Portage Island sites with the exception of N8 and N10. Overall, the sites dominated by marine-estuarine habitat had greater biomass in comparison to predominately fresh water sites.
131
Macroinvertebrate Biomass
1.4
1.2
1
0.8 Phase 1 Phase 2 Phase 3 0.6 Total Biomass (g)
0.4
0.2
0 L0 L1 L2 L3 L4 L5 L6 L7 L8 L10 P1 P2 P3 N1 N2 N3 N4 N5 N6 N7 N8 N9 N10 N11 Sampling Site ID
Figure 74. Total dry weight biomass of macroinvertebrates collected during three different phases (in 2004) at 24 sampling sites in the Lummi River (L), Nooksack River (N), and Portage Island / Puget Sound (P). From Ross and Weispfenning (2000).
To attempt to reveal a possible relationship between macroinvertebrate biomass and substrate type, an average substrate complexity index value was calculated for each of the 24 sampling sites. We were testing the hypotheses that vegetated substrate was more complex and would sustain greater benthic macroinvertebrate biomass than non- vegetated substrate. The index values assigned to each substrate category are as follows: 1 = silt and mud, 2 = sand, 3 = gravel and shell, and 4 = substrate with vegetation and organic debris. However, a Pearson’s correlation between total macroinvertebrate biomass and substrate complexity of the 24 sampling sites did not reveal a significant relationship based upon our substrate complexity index values (r = 0.104 with correction) (Ross and Weispfenning 2004).
132
Table 6. Macroinvertebrate dredge-sampling sites and their rank in each of three categories: species abundance, biomass, and diversity. The rank of 1 is highest. The top 5 ranked positions are highlighted in red.
Species Species Site Abundance Biomass Diversity ID Habitat Type Channel Type Substrate Rank Rank Rank
L4 Agricultural Floodplain Relict Tidal Mud, wood debris 1 2 12
L6 Sand Flat Tributary Mud, fine sand 3 3 7
L3 Agricultural Floodplain Tributary Mud, silt 4 8 18
L0 Agricultural Floodplain Tributary Mud, silt 9 5 11
L2 Agricultural Floodplain Tributary Mud, silt 15 22 19
N5 Forested Floodplain Tributary Coarse sand 21 17 23
N3 Scrub-Shrub Tributary Mud, silt 19 16 13
L5 Mud Flat n/a Sand, shell debris 12 10 1
P2 Mud Flat/Eelgrass n/a Mud, eelgrass 6 7 2
L7 Nearshore n/a Fine sand 2 1 9
P1 Sand Flat n/a Sand, gravel, eelgrass 14 4 3
L8 Sand Flat n/a Fine sand, mud 11 13 4
L10 Sand Flat n/a Sand, some mud 7 9 5
P3 Nearshore n/a Sand, gravel, eelgrass 5 6 6
N11 Sand Flat n/a Sand, gravel 8 20 8
N10 Sand Flat n/a Coarse sand 13 12 15
N8 Salt Marsh Blind Mud, wood debris 10 11 10
N1 Agricultural Floodplain Mainstem Coarse sand 18 21 17
N6 Scrub-Shrub Mainstem Sand 17 15 20
L1 Agricultural Floodplain Distributary Mud, silt 16 18 16
N4 Forested Floodplain Distributary Sand, silt, wood debris 20 14 14
N2 Scrub-Shrub Distributary Sand, silt 22 23 21
N9 Salt Marsh Distributary Medium sand 23 19 24
N7 Forested Floodplain Distributary Sand, wood debris 24 24 22
133
Species Abundance Following is a summary of relative abundance of macroinvertebrates sampled during the three stages. The numbers of species and individuals sampled are described in Appendix A, tables 12, 13, and 14. There is variability between total abundance and abundance per species between the two sampling years. We have not developed hypotheses to explain this variability.
Phase 1 Samples The greatest number of organisms was collected during Phase 1 sampling. Over 5,500 organisms were collected between February and March in 2003; over 5,200 organisms in 2004. Chironomids and Corophium sp. were most abundant. This is encouraging, due to the known importance of these invertebrates as food resources to salmon in the juvenile life stage (Gray et al. 2002, Cordell et al. 1999).
Benthic samples at most of the 24 sampling sites consisted primarily of annelids in both 2003 and 2004. In addition to annelids, samples collected from Nooksack Delta sites during Phase 1 in 2003 yielded mostly amphipods and chironomid larvae. Lummi Delta sites produced large numbers of amphipods, and Portage Bay sites also had large numbers of decapods. In 2004, Corophium sp. percent abundance was high only at L4 and N7, while leptochelia was only abundant at L8. Eogammarus was most abundant at L7 making up 20% of the sample. The percent of chironomids per sample were fairly low across all sampling sites except for N6 and N8. Copepods were essentially nonexistent from all sample sites; however, N10 and N11 samples consisted of approximately 60% and 80% copepods respectively.
Phase 2 Samples The total number of invertebrate organisms collected during Phase 2 (April-May) in the estuary and nearshore was significantly lower than the number collected in Phase 1 (slightly less than 3,000 in 2003, and just under 2,000 in 2004).
In 2003, annelids were the primary species present. Chironomids and amphipods dominated Nooksack Delta and Lummi Delta sites in Phase 2, and gastropods were of secondary abundance, to annelids, in the Portage Bay sites. In 2004, annelids were abundant in the Lummi River and Portage Bay samples. Chironomids dominated the Nooksack River sites N1 through N4 and were limited to these four sites and L0. The benthic sample from N10 consisted of 100% Corophium sp., which was also fairly abundant at L6 and N8. Eogammarus made up 30% of the L10 sample but it was less than 10% at L5, L6, L8, and N8. Approximately 50% of the L8 sample was comprised of leptochelia, and was not present in any of the Nooksack River sites. Copepods were not found within any of the samples during phase 2 surveys.
Phase 3 Samples Phase 3 had slightly more invertebrate organisms collected than Phase 2, but considerably less than the total collected in Phase 1. All samples collected during Phase 3 were accounted for and contributed to the total. The total number of organisms collected in Phase 3 is less than half of that collected in Phase 1 in 2003 and 2004. Phase
134
3 of invertebrate collection coincides with the bulk of juvenile salmon migrating through the estuary and nearshore.
Annelids and chironomids were the most abundant macroinvertebrates sampled overall. Corophium sp. made up 100% of the sample collected at N9 and was approximately 60% of the N8 sample. Eogammarus was only present at L5, N10, and P1. Copepods were not found within any of the samples collected during phase 3 surveys.
Species Diversity Since macroinvertebate species hatch and dwell in aquatic habitats at different times, the diversity of macroinvertebrate species presence at a site is an indicator of the temporal continuity of prey available to juvenile salmon. In addition, macroinvertebrate species diversity is generally recognized as an indicator of environmental quality in aquatic habitats.
In the LNR study, taxonomic family-level diversity was used to classify the sample results by species rather than species-level diversity because not all macroinvertebrates collected were identified to the species level. Species diversity was calculated with the Shannon-Weiner Index (H’) and ranked by site (Figure 76). In 2004, the number of macroinvertebrate families across all sampling sites was highest during Phase 3 sampling with 31 families observed. Phase1 had a total of 30 families, while Phase 2 had 23 families. Portage Bay sites had the highest average number of macroinvertebrate families per sample (Phase 1 mean = 17 families, Phase 2 mean = 11 families, Phase 3 mean = 19 families) followed by Lummi River sites (Phase 1 mean = 10 families, Phase 2 mean = 10 families, Phase 3 mean = 9 families) and Nooksack River sites (Phase 1 mean = 3 families, Phase 2 mean = 2 families, Phase 3 mean = 3 families).
Phase 1 Samples Taxonomic richness calculated for sites sampled in 2003 was highest in the Portage Bay and lower Lummi Delta sites. Mainstem Nooksack River sites were low overall, with the exception of Silver Creek and Kwina Slough (high), and the Blind Channel (moderate). In 2004, diversity of macroinvertebrate families was greatest at L7 (H' = 2.17). Lummi River and Portage Island sites on average had a greater number of families represented in the benthic samples than the Nooksack River sites. Of the Nooksack River sites, the highest number of macroinvertebrate families were found in three sites with significant marine influence: the Blind Channel, and the eastern and western delta nearshore areas (N10 and N11, respectively).
Phase 2 Samples In 2003, diversity was highest in Portage Bay, followed closely by the Lummi Delta sites. Nooksack River sites had the lowest diversity on average, but the East Channel mainstem site faired high. In 2004, macroinvertebrate diversity was highest at L5 (H' = 2.35). Family-level diversity was greatest at the Lummi River and Portage Bay sites. There was not much variation in the number of families between the Nooksack River sites during Phase 2 sampling.
135
Phase 3 Samples In 2003, diversity was highest at L0, and in 2004, family diversity was greatest at L5 (H' 2.61). Lummi River and Portage Island sample sites tended to have the highest number of macroinvertebrate families. Nooksack River sites sampled in 2003 had low overall diversity, but sites N3 (Silver Creek) and N8 (blind channel) were relatively high. In 2004, macroinvertebrate family diversity for the Nooksack River sites was similar to richness calculated in Phase 1, highest at N8, N10, and N11.
Macroinvertebrate Family Diversity
25
20
15
Phase 1 Phase 2 Phase 3 # of families 10
5
0 L0 L1 L2 L3 L4 L5 L6 L7 L8 L10 P1 P2 P3 N1 N2 N3 N4 N5 N6 N7 N8 N9 N10 N11 Survey Site ID
Figure 75. Diversity of macroinvertebrate families present in benthic samples collected during three phases.
Summary of Macroinvertebrate Sampling Phase 1 sampling consistently produced the most organisms per sample, followed by Phase 3, and Phase 2. Phase 1 sampling accounted for over 50% of the total abundance both years sampled, and Phase 2 and Phase 3 each accounted for around 25%. This high winter yield of invertebrates coincides with aquatic vegetation (eelgrass, kelp, algae) dieback and decomposition in the fall, and detrital buildup in the winter. Detritus feeds invertebrates, and may support healthy populations in the estuary and nearshore habitats that maintain vegetation communities. River sampling sites are characterized by abundant wood along streambanks and at the front of the delta. These wood assemblages accumulate detritus, in turn attracting invertebrates. However, low invertebrate sample sizes found in river sites may be a product of sampling gear limitations in high-energy
136
environments. It is evident that marine-influenced sites in Lummi and Portage Bays, and the nearshore sites were the most productive and diverse.
Overall, the diversity and biomass results between 2003 and 2004 were not significantly different. Results from both years show that sites with brackish or saline water quality yielded the greatest biomass. The diversity of macroinvertebrate families was highest at the sampling sites directly influenced by the marine environment. Of the Nooksack River sites, N8, N10, and N11 are the most highly influenced by marine waters and consequently had the highest combined family diversity and biomass, compared to all river sites. Sites L5, P2, P3, and L7 in Lummi and Portage Bays, and the nearshore sites yielded the highest diversity of macroinvertebrate populations overall. These trends were expected since estuarine environments are extremely productive systems as a result of nutrients made available from freshwater inputs and oceanic upwelling, and of the diversity of primary producers.
There appears to be a seasonal / temperature effect on benthic macroinvertebrate diversity between the sites sampled. Differences in life history characteristics such as temperature and light tolerance, reproduction, and behavior may favor some species in the winter and others in the spring, while microhabitat differences between the sampling sites may also likely affect species diversity and abundance (Ross and Weispfenning 2004).
Juvenile salmonid arrival to the estuary varies temporally by species. During Phase 1 of macroinvertebrate collection, the period with highest invertebrate abundance, the greatest number of salmonid species is present in the estuary (Ross and Weispfenning 2004, MacKay 2004, in prep). January and February mark the arrival of chum, pink, coho and early chinook fry migrants to the estuary. Early salmon arrivals to the estuary are presumed to reside in the delta and nearshore sites, feeding on diverse populations of invertebrates. During Phase 3, the bulk of chinook juveniles are entering the estuary, residing here before dispersal to nearshore sites. Invertebrate abundance during this time is notable.
The data collected in this assessment activity tends to support the conclusion that the macroinvertebrate population available to juvenile salmonids in the Nooksack estuary and nearshore is diverse (Table 5). Although abundances varied by season and site, the overall species richness of salmon prey items was high. Samples collected bi-weekly contained both larval and adult insects. The samples also contained vertebrate food sources including herring eggs, and larval prey fishes such as sand lance, herring, and surf smelt.
Substrate size and water velocity were two notable variables at each site that produced differences in invertebrate diversity and abundance. Sites with high velocity and few organisms found in the samples included the mainstem Nooksack River, swift distributary channels off of the mainstem, and Kwina Slough. Correlation of species richness and abundance to substrate size at each site revealed that channels with lower flows and subsequent fine grain substrates such as fine sand, silt, and mud were able to produce a more diverse array of salmonid prey items such as copepods, ostracods, and
137
mysiids. High-energy habitats feature coarser unstable substrates that are not conducive to retaining the abundance of lower level food web organisms. Table 5 above summarizes our findings on the abundance and diversity of macroinvertebrate populations sampled in the assessment in relationship to channel types, substrate composition, and the terrestrial habitat type the channel is flowing through.
Fish Usage Although salmonids are known to migrate through estuaries, surprisingly little is known about their utilization of these habitats. Many researchers in the Pacific Northwest maintain that estuaries serve as more than migration corridors (Reimers 1973, Levings 1982, Miller and Simenstad 1997), suggesting that estuaries are cornerstone habitats of the salmonid life history since they are utilized when physiological adaptation, foraging, and predator avoidance are critical (Healey 1982, Simenstad 1982 cited in Bush 2003).
Pacific salmon use the estuary for transition between life history stages twice during their life cycles. The first transition is between freshwater smolt and saltwater smolt in the juvenile life stage. The second is between sexually mature saltwater adult and spawning freshwater adult life stage.
It is important to note that salmon require different resources of the estuary during their migration through it. In spite of these differences, there are similarities between estuary use by juveniles and use by adults. Adults, like their juvenile counterparts, use the estuary for physiological transitioning between marine and fresh water habitats. They seek refuge from predators; their primary hunters include marine mammals, humans, and birds. Habitat complexity is an important aspect of predator refuge; attributes such as overhanging vegetation, undercut banks, and wood assemblages all contribute to essential estuarine hiding habitat. Passage barriers create problems for both adult and juvenile salmon, preventing them from maximizing habitat potential.
Estuaries provide a range of habitats for juvenile salmon. Smoltification, feeding, and predator avoidance are primary functions provided by estuary habitats. Smoltification is the process that bridges the freshwater fry migrant life stage to the saltwater adult life stage. It encompasses the physiological changes necessary for juvenile salmon to adapt to brackish and full-strength salt water. The smoltification process begins with the downstream migration from the upper freshwater habitat of a salmon’s natal stream, and ends with the final transformation in seawater. Length of time needed for smoltification depends on the salmon species, and can occur within 30 days of emergence (pink and chum fry), up to one or two years after emergence (coho, sockeye and chinook yearlings). What determines whether fish will hold and rear in the river, or migrate immediately downstream to the estuary is unknown.
Young salmon in the smolt life stage undergo a morphological, physiological, and behavioral metamorphosis that prepares them to life in seawater. During this period, their camouflage parr marks disappear, and they turn silver for ocean living. Their osmoregulatory mechanisms begin adjustments that will allow them to process salt water for survival. They cease territorial behavior and form schools as they begin their seaward
138
journey. Just prior to the smolt stage, the endocrine system undergoes major transitions; thought to be induced by thyroid hormone activity (Hasler and Scholz 1988). This life stage is of critical importance to the homing process that brings adults back to their natal streams. It is during this smoltification period that salmon “imprint” to some property of their natal tributary that serves later to identify it when they return to spawn (Hasler and Scholz 1988).
Estuary habitat must provide the attributes necessary to facilitate the completion of smoltification. Water quality is highly important. Salinity exposure should be gradual. Saline water penetrating the estuary in the form of a dense water mass known as the salt wedge should be accessible to young salmon in habitats that provide cover. Water temperature needs to be hospitable to both the salmon and their prey items. Areas exposed to direct sunlight, such as tide flats may not be as hospitable to young salmon as those that can provide cover, especially when the water is not turbid. Salt marsh, scrub- shrub, and forested habitats within the estuary are ideal saline transition habitats. Tidal channels in salt marsh habitat, as well as wood-armored and undercut banks in scrub- shrub and forested habitats can also provide necessary cover.
Feeding is the other major objective of juvenile salmon during smoltification. Feeding and growth of juvenile salmonids, particularly chinook salmon, are linearly related. Successful feeding supports rapid growth, which increases survival due to an increased ability to avoid predators. This ability is a result of faster swimming speeds in larger individuals. A larger size allows the transition to a new variety of prey resources in new habitats (Healey 1998, Kerwin and Nelson 2002). The feeding and growth of juvenile chinook are also functions of fish size and habitats occupied. Kerwin and Nelson (2002) note that the diet of salmonid fry is dominated by insects. Fingerlings feed on insects in freshwater channels and benthic invertebrates such as amphipods and corophium in the lower estuary. Research by Healey (1998) found that salmonid growth is typically higher in estuarine habitats than in freshwater habitats (cited in Kerwin and Nelson 2002).
Once out of the natal estuary, juveniles may either migrate directly to the open ocean or migrate through nearshore habitats to other estuaries and where they continue to feed and prepare for ocean conditions. Brennan et al (2004) found juvenile chinook in Puget Sound nearshore habitats year round, suggesting that these fish exhibit opportunistic behavior and leave this habitat only when ready. Levy and Northcote (1981) observed a twice-daily pattern of migration from low-tide refuges to the brackish and fresh water marsh areas and back again, continuing throughout the period of residence of fry in the estuary. As the estuarine residence period progressed, the authors found that higher concentrations of young fish moved seaward through the delta area. This is believed to be partly due to larger fish preferring deeper water and partly to allow them to osmoregulate in higher salinities (Healey 1998). High water temperatures in shallow delta areas, especially later in the season, likely accelerate movement toward nearshore environments.
In the Nooksack estuary and associated nearshore environments, extensive sampling was conducted to characterize juvenile salmonid use of estuarine habitats. Information on
139
fish utilization presented here is from Nooksack River screwtrap and estuary beach seine studies conducted in 2003 and 2004 by the Lummi Natural Resources Department. While these projects provide some observations of juvenile salmon in the estuary, they may not reflect historic distributions nor do they cover all habitats in the estuary that may be of importance to salmon. A more detailed description of these studies can be found in the following project reports: (MacKay, 2004a and 2004b; and Pfundt, 2004a & 2004b).
The sampling effort included 46 beach seine sites and a river screwtrap (Figure 77). The sites were distributed across 6 geographic areas in 6 different habitat types: protected nearshore (Bellingham Bay, Hale Passage, Portage Island), exposed nearshore (Strait of Georgia), delta nearshore (Nooksack River, Lummi River), salt marsh, scrub-shrub, and forested floodplain. The geographic designations were assigned to characterize the migratory pathways of juvenile salmon. The river screwtrap is located in the lower mainstem of the Nooksack River, approximately 5.8 miles upstream of the mouth at the upper extent of the estuary.
To make comparisons of fish abundance over time between the seine sets and the river screwtrap, the data need to be related to the sampling effort. The river screwtrap uses hours as a measure of sampling effort. The hours fished varied by month. Abundance over time for this device is indicated by catch per hour. For the beach seine study, average catch per seine set was calculated and used as a measure of abundance. Sampling at all beach seine sites occurred bi-weekly. In 2004, two nets were used of different lengths (40-foot and 60-foot), but with the same depth (10’) and mesh size (1/8”).
As with any sampling device, there are inherent limitations that introduce sampling bias. The river trap is limited to a single location of eight feet in width, over the deepest part of the river channel, to a maximum depth of 4 feet. The beach seine is used only along beaches and stream margins, not in the open water column. Both devices are more efficient at capturing salmonids when the water is turbid and gear avoidance is reduced. Water turbidity in the river and nearby marine areas can vary greatly in a few hours following storm events; this is likely a significant factor affecting catches and causing variation in the data. The beach seine can only be used in areas clear of snags, which excludes many woody areas of the Nooksack Delta, used extensively for cover by rearing juvenile chinook and coho salmon (Mossup and Bradford 2004, Hicks et al. 1991, Dunphy pers. comm.). Beach seines are also not effective at sampling shallow gradient mud/sand flats, or rocky marine shorelines. Neither device can provide us with information on juvenile salmon use in the vast marine sub-tidal areas.
140
Screwtrap
Figure 76. Nooksack Estuary study area showing geographic study areas and delta habitat zones.
141
In 2004, fish were sampled for a longer time period than in 2003. As a result of this refinement, the 2004 data are more suitable for presenting temporal patterns of abundance. In 2003, more locations (90 sites) were sampled and overall catches were higher for chinook.
Hatchery origins of sampled chinook, coho and steelhead were detected by presence or absence of an adipose fin clip or coded wire tag. Unfortunately, not all hatchery chinook had the fin clip mark; therefore, hatchery chinook may be under-reported in our data. This significant problem, which limits our understanding of non-hatchery chinook, resulted from the absence of a detectable mark on approximately half of the Kendall Creek North Fork chinook.
Chinook Salmon At present, there are three chinook stocks in the Nooksack River recognized by Tribal, State, and Federal Agencies (WDFW 1993). Adult escapement and spawn timing are used to describe the stocks. Two are early-timed indigenous stocks and one is a fall- timed stock of hatchery origin. The North Fork chinook and South Fork chinook stocks are early return (spring-run) chinook that reside in the Nooksack North and South Forks, respectively. These stocks are listed as protected resources by the National Marine Fisheries Service (NMFS). All three stocks overlap to some degree in spawning location and timing. Generally, the North Fork chinook has a peak spawn period in late August. The South Fork stock peak occurs about two weeks later than the North Fork group and the Samish Fall stock peaks near the middle of October.
South Fork chinook are considered indigenous and spawn primarily in the South Fork mainstem and larger tributaries. NMFS describes them as essential for recovery of the threatened Puget Sound Evolutionary Significant Unit (ESU). As a native stock, they have a unique genetic background that reflects their life history as characterized by unique spawn timing and geographic separation.
The North Fork Nooksack Fall stock, although not described as threatened, is a significant population in the Puget Sound ESU. This is a supplemented stock propagated from non-hatchery-origin North Fork native chinook. This stock is sustained by the collection of broodstock from the WDFW Kendall Creek Hatchery. Its spawning range is generally the mainstem North Fork and larger associated tributaries, including the Middle Fork.
The third stock present in the Nooksack River is the Samish River Fall chinook population. It is a late-timed hatchery origin stock that has existed for several decades in the Nooksack River. The genetic origins of this stock are primarily the Green River, Soos Creek, and the Samish River. Samish Fall run chinook spawn in the mainstem Nooksack River above approximately river mile 10.
The three chinook stocks described above produce juveniles that migrate through the estuary, albeit at different times, depending on life history strategies. At the time of emergence, there is an extensive downstream dispersal of chinook fry (Healey 1998).
142
Some fry take up residence near their natal nest, others begin the downstream migration toward the estuary. Once started downstream, chinook fry may continue migrating downstream to the river estuary, or stop migrating and take up residence in the stream for a period of time. A large downstream movement of chinook fry immediately after emergence is typical of most populations (Healey 1998). Chinook fry can spend anywhere from several days to a year in freshwater prior to migrating to the estuary. Such variability can occur within a single stock of chinook, but more typically a single stock would be classified as either ‘ocean-type’ (fry) or ‘stream-type’ (yearling) chinook, the latter representing those fish that spend one year in freshwater (Kerwin and Nelson 2002). Other terms used to describe these two life history strategies include ‘yearling’ and ‘sub-yearling’ chinook. Chinook salmon arrive to the estuary as one of two types. Fry migrants, those that arrive in the estuary shortly after emergence, feed heavily here, and rear to nearly double in size before leaving. Yearling migrants, those that arrive after rearing in fresh water for nearly a year, rely less on the estuary for growth, and have been observed to migrate directly to brackish nearshore habitats to complete the smolt life stage (Healey 1998).
The outmigration period for specific stocks of Nooksack chinook is not well known, although the primary run occurs from January through July. Juvenile chinook have been found in the delta and nearshore areas of the estuary from July until December but are much less common during that time than in the spring and early summer. Estuary residence times for Nooksack River chinook is not known at present but can be determined from mark and recapture studies, if they are incorporated in future sampling methods.
Estuarine residence times of chinook vary by arrival time, water temperature, streamflow, and fish size. Chinook are known to reside in the estuary for a number of days or months, depending on the aforementioned variables. The importance of estuarine rearing on chinook production has been determined by scale analysis of returning adults to an Oregon estuary; the survival of fish that remained in the estuary longer was greater than that of migrants that left the estuary early (Reimers 1973).
Juvenile diets vary considerably from estuary to estuary and from place to place within an estuary. Chinook generally cohabit with other salmonids in estuaries, especially with chum salmon. Although they often eat the same organisms, the correlation between their diets was found to be weak in the Fraser and Nanaimo River estuaries (Sibert and Kask 1978). They also found that in the Nanaimo River estuary, the chinook diet correlated poorly with the diet of cohabiting coho, and was more similar to the diets of some non- salmonid cohabiters such as herring (Clupea pallasi), sticklebacks (Gasterosteidae spp.), shiner perch (Cymatogaster aggregate), and sand lance (Ammodytes hexapterus). Research by Dunford (1975) found that chinook were more efficient predators of chironomid larvae than their chum rivals, and were able to capture and eat Neomysis that chum could not.
Seasonal changes in diet reflect seasonal changes in the abundance of prey items. Levy and Northcote (1981) reported that chironomid larvae and pupae were the most important
143
diet items of ocean-type chinook in tidal channels throughout Fraser River marshes. Of secondary significance were Daphnia, Eogammarus, Corophium, and Neomysis.
New chinook salmonid recruits to the estuary tend to prey on larval and adult insects, and various amphipods (Healey 1998). Research by Simenstad et al. (2003) found that Daphnia spp. and other zooplankton comprised much of the diet of juvenile chinook on the brink of leaving their freshwater habitats for Puget Sound. In nearshore areas, insects, epibenthic crustaceans, and polychaete annelids were prominent. Koehler et al. (2000) found that juvenile chinook in the littoral zone of the Salmon Bay estuary near Lake Washington in Seattle fed primarily on aquatically-derived insects (59% of diet biomass); zooplankton accounted for 27% of their diet biomass; and 5% of the juvenile diet biomass consisted of terrestrially-derived insects. In another study, Brennan et al. (2004) analyzed the stomach contents of juvenile chinook salmon seined from Puget Sound nearshore habitats and found that in over 800 fish caught over two sampling seasons, 50% of the diets were terrestrial riparian-derived insects, nearly 30% were marine planktonic, nearly 20% marine benthic invertebrates (primarily annelid worms), and the remainder consisted of aquatic vegetation.
Larger-sized juvenile chinook in the estuary are known also to feed on chum salmon and pink salmon juveniles, as well as larval-stage herring, sand lance, and longfin smelt (Spirinchus thaleichthys) (Hart 1980). This happens later in the season, as they become larger, more effective predators. Juvenile chinook salmon are found in nearshore environments year-round (Brennan and Higgins 2004), but concentrate in areas with abundant prey. They may migrate between estuaries using the nearshore as a corridor, as they feed and grow on their way out to sea. Figure 77 represents a generalized view of juvenile chinook feeding trends in Puget Sound estuaries. It describes the diverse, opportunistic feeding behavior exhibited by these fish.
144
Figure 77. An example of temporal changes in the diet of juvenile chinook salmon in the estuary. At the top of the figure, [O] refers to other diet items (from: Groot and Margolis, 1998).
There are four life history strategies that have been identified for Puget Sound chinook juvenile migrants (Averill et al. 2004): fry migrants, delta fry, parr migrant fingerlings, and yearlings. These strategies are delineated by age upon arrival and residence time in the estuary. Due to gaps in Nooksack estuarine residence data, it is difficult to determine which of the ocean-type strategies are common in the Nooksack estuary. The fourth strategy, exhibited by stream-type chinook juveniles (yearlings), is easier to identify, as they are considerably larger than their fry counterparts. Estuarine residence of these fish cannot be estimated with any degree of certainty, again, due to gaps in sampling data. As a result, Nooksack chinook juveniles in the estuary are referred to as either fry or yearling, based on size.
Estuarine life stage requirements for fry and yearling chinook differ in several ways. Food resource needs are pertinent for both; however, fry chinook require smaller-sized prey items than those that can be assimilated by larger yearling chinook. Feeding opportunities for yearling chinook are better than for their smaller fry counterparts. In the estuary, fry rely on small detritivores, shellfish larvae, and soft-bodied items like annelids. Yearlings may also feed on these items, but are capable of additionally preying on larger items such as appropriately sized fish, drift insects, and large larval stage invertebrates.
145
Swimming speed and strength differ between the two life history strategies, as well. Fry migrate slower through the estuary than larger juveniles, a characteristic attributed to their preference of slower velocity streambank areas. Their orientation to flow direction also affects their migratory patterns; smaller fish face upstream during migration, whereas larger juveniles usually face downstream as they navigate channels (Schaffter 1980, cited in Allen and Hassler 1986). Yearling chinook can better navigate higher flow velocities than their smaller fry counterparts, and may have more control of their distribution within estuarine habitats. This skill may assist yearling fish in more efficient predator avoidance, allowing them to swim away from threats faster than fry juveniles. High flows may force fry out of the delta earlier than necessary, and may limit their residence time in the estuary. Low flow refugia within the estuary, such as areas along channel margins, within log assemblages, or in the pools scoured out beneath them, may be more critical to fry than to yearling chinook, given the yearling’s improved ability to navigate higher flows. Areas in the estuary that cater to the fry migrant’s hydrologic needs include the margins of small side and distributary channels, and blind channels.
Fry migrants have more potential predators than their larger yearling counterparts and have a greater need for protective cover from their predators. Undercut bank habitat and overhanging vegetation in scrub shrub and forested wetland landscapes provides protection.
Salinity tolerance increases with the size and rate of growth in chinook (Allen and Hassler 1986). Fry-sized chinook have been observed to prefer lower salinity water during estuarine the rearing period, larger fish are better acclimated to tolerate sharp salinity gradients (Healey 1982). Brett (1952) estimates salinity requirements for rearing juvenile chinook salmon between 12-13 ppt. Chinook juveniles are also more tolerant of higher water temperatures than other Pacific salmon; optimum rearing temperatures are between 12-14°C (Brett 1952). Given the advantages that larger fish have for making the best use of estuarine habitat, we can speculate that yearling migrants are better suited for surviving in the estuary and migrating to marine habitat than fry.
Although larger chinook juveniles are more efficient navigators of high discharge conditions, fry and yearlings both prefer surface waters in shallow flats and deepwater channels (Allen and Hassler 1986). The affinity of juvenile chinook for deep pools prevails in fresh as well as estuarine waters; Roper et al (1994) concluded that fry migrants were strongly associated with pools in estuarine habitats, and Glova and Duncan (1985) found that juvenile chinook prefer deep reaches of intertidal and estuarine habitats (McNeil 2001). Levy and Northcote (1981) researched the relationship between occurrence and abundance of chinook fry in various marsh habitats according to the physical characteristics of the habitat. Their results suggest that young chinook prefer tidal channels with low banks and many low tide refugia (wood, vegetation). Chinook tended to be associated with larger tidal channels with high complexity that provided diverse microhabitats (McNeil 2001).
146
Fish sampling efforts by Lummi Natural Resources delineated juvenile chinook in the estuary by fry (0-age) and yearling (1+ age) individuals. We made this distinction based on fish size using nine years of catch data from scale samples taken of fish caught at the river screwtrap. Subsequent fish sampling efforts in the estuary could not determine residency time. We could only document that at a given time and site, fish were present or they were not.
Fry Migrants These fish represent the majority of chinook juveniles migrating through the Nooksack estuary. They arrive as early as December and January, peak in May, and continue migrating through June and July. Fry entering Nooksack estuary and nearshore habitats are a combination of wild and hatchery stocks. Hatchery releases occur in the beginning of May; these fish arrive in the estuary shortly thereafter. Due to past inconsistencies in hatchery marking protocol, not all fish are released with a mark, and determination of origin is difficult. The temporal variation among the fry migrant population does not create significant differences in their estuarine requirements, but coincides with a distinct shift in resource availability and habitat variables. Water temperatures in delta channels disconnected from the river begin to increase in May, some nearing sub-lethal limits, and we observed a slight decrease in benthic invertebrate populations between March and June. Conversely, shelter opportunities improve as riparian vegetation produces leaves and flowers, drift insects populations increase, and kelp revegetates into thick beds used by juvenile salmon for cover.
Yearling migrants Juveniles that enter the estuary after rearing for a year or more in freshwater habitats are described as yearling migrants. These fish typically enter the estuary at a fork length between 80 – 120+ mm (Aitkin 1998), and spend a short time in the estuary before moving out to the nearshore.
Yearling outmigrants are not common in the Nooksack system. From river and beach seine collections a total of 28 yearlings were caught in 2003, compared to 86 in 2004 (Figure 78). These hauls were 0.3% of the total chinook catch in 2003, and 1.4% of the total in 2004. The ratio of hatchery yearling to river-origin yearling chinook present in the estuary is difficult to determine, considering the inconsistency in hatchery marking practices prior to analysis of these data.
There seem to be two periods when yearling chinook arrive in the estuary, smaller numbers early in the season, and greater numbers in the mid-season. Most yearlings were caught during the middle of the outmigration season; however, in 2004, there appeared to be an initial period of yearling outmigration in January and early February. These catch results, when compared to fry migrant numbers, may underestimate their abundance in the estuary, in part due to the faster swimming speed of these larger fish and their ability to detect and avoid sampling gear, especially when the water is clear.
147
Yearling Chinook - Total Trap and Seine Catch, 2003 & 2004 40
35 2003
30 2004 25
20
15
10
5
0 1/17 1/31 2/14 2/28 3/13 3/27 4/10 4/24 5/8 5/22 6/5 6/19 7/3 7/17
Figure 78. Total Nooksack Estuary bi-weekly catch of yearling chinook in the river screwtrap and in beach seines in 2003 & 2004.
Total chinook catches in the estuary are described in tables 7 and 8, below. Screwtrap data represents new Nooksack River stock arrivals to the estuary; beach seine data may include fish from other systems using the Nooksack estuary and nearshore for rearing. It is important to note that catch efforts varied somewhat between 2003 and 2004; these data are presented to show relative numbers of chinook sampled in the estuary during the outmigration season.
Table 7. Chinook catch in the river trap and at beach seine sites in the Nooksack delta and nearshore areas in 2003. Year Gear Area Marked Unmarked Yearling Total Chinook Chinook Chinook Chinook Fry Fry 2003 Screwtrap River 2,120 5,615 10 7,735 Beach Seine Delta 79 1,528 8 1,607 Beach Seine Nearshore 401 395 10 796 Total 2,600 7,538 28 10,138
148
Table 8. Chinook catch in the river trap and at beach seine sites in the Nooksack delta and nearshore areas in 2004.
Year Gear Area Marked Unmarked Yearling Total Chinook Chinook Chinook Chinook Fry Fry 2004 Screwtrap River 2,523 2,494 53 5,017 Beach Seine Delta 186 336 13 522 Beach Seine Nearshore 320 82 0 401 Total 3,028 2,912 66 5,940
The timing and size distribution of all juvenile chinook entering the estuary, based on trap catch records, is described in figures 79 and 80, below. These data suggest that earlier arrivals to the estuary are smaller but more plentiful. Later arrivals are larger in size, due in part to an increase in feeding opportunities between the two periods measured, and the presence of yearling juveniles in catches.
Fork length of chinook arriving to the estuary, April 60 50 40 30 20
Numbers of chinook 10 0 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 Fork Length (mm)
Figure 79. Number and size of early migrant chinook to the Nooksack estuary.
149
Fork length of chinook arriving to the estuary, June 20
15
10
5 Numbers of chinook 0 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 Fork Length (mm)
Figure 80. Number and size of later arriving migrant chinook to the Nooksack estuary.
There is not enough data at this time to delineate variation in estuarine habitat utilization, or to predict habitat preferences of chinook. We do know that the total chinook catch was low in both years, until the onset of hatchery arrivals in May. Hatchery-marked chinook sampled in beach seines set in the Nooksack Delta and surrounding nearshore sites declined somewhat sooner than their numbers did in the river trap, possibly due to later fish exhibiting a more rapid movement through the estuary and into nearshore sites.
The optimal arrival time of juvenile chinook to the Nooksack estuary to accommodate their estuarine habitat requirements is between late spring and early summer. Water temperatures and salinities are maintained within the preferred range, and spring precipitation and snow pack runoff maintain adequate flows for navigation and maximum use of channel habitats. We have observed the bulk of chinook migrants arriving to the estuary during the late spring and early summer period (Figures 79 and 80, respectively).
In conclusion, juvenile chinook in Nooksack estuary and nearshore habitats are primarily fry migrants. Considering the existence of habitat that provides juvenile salmon with food and shelter resources, we hypothesize that salmon are rearing in the estuary; however, the degree to which these migrants may saturate existing habitat is unclear. Effective determination of chinook residency in estuarine habitats will likely require a combination of mark and recapture research and improved gear efficiency. The advantage of our current sampling regime with the ability to capture migrants in the screwtrap as they arrive to the estuary could facilitate mark and recapture analysis. Temporal documentation of the presence of these marked fish elsewhere in the estuary using beach seines would be possible. The results would support the determination of residency times and further describe the specific needs of juveniles in the estuary. Beach seines used to sample fish further downstream in the estuary are limited by their ability to
150
effectively sample wood complexes and other high flow and predator refugia that may be used by juvenile salmon. The possibility of the net snagging on wood prevents successful sampling of habitats that maintain high volumes of wood, such as side and distributary channels with significant riparian cover and instream woody debris, and the upper intertidal zone during high tide events.
Coho Salmon Coho salmon (O. kisutch) are known to rear in their natal freshwater habitats for over a year after emergence, often up to eighteen months after. With moderate water temperatures and an abundant food supply, coho fry will grow from 30 mm at emergence in March to 60–70 mm in September, to 80–95 mm by March of their second year, and to 100–130 mm by May (Sandercock 1998; Rounsefell and Kelez 1940). Water temperatures between 12-14°C are optimum for maximum growth efficiency (Bjornn and Reiser 1991, Brett 1952). In some river systems, coho may stay two, three, or even four years in fresh water before outmigrating; however, most Nooksack river migrants hit the estuary during their second year (Pfundt, pers. comm. 2004; MacKay 2000).
The size of fish, flow conditions, water temperature, dissolved oxygen conditions, day length, and food availability all affect the exact time of migration (Shapovalov and Taft, 1954). In a single river system, there are year-to-year variations in the timing of coho smolt migration, related to environmental factors. Smolt trapping efforts at the mouth of the Nooksack River by Lummi Natural Resources staff between 1994 and 2003 reveal a consistent pattern of coho migration to the estuary between the first week of May and the last week of July (MacKay 2000, 2004 in prep.).
In brackish and salt water, feeding by coho salmon juveniles is active, and growth rapid. Young fish remain near the surface, feeding on herring larvae and sand lance. Near the Nooksack estuary, the coho salmon’s estuarine diet is based mainly on small fishes such as the aforementioned herring and sand lance, and kelp greenling (Hexagrammos decagrammus), rockfish (Scorpaenidae spp.), and eulachon (Thaleichthys pacificus). Other important species in the diet of juvenile coho salmon include crustaceans such as copepods, amphipods, and barnacle and crab larvae (Hart 1980).
Like chinook, coho salmon produce both zero-age and yearling outmigrants. Unlike chinook, most coho in the Nooksack River outmigrate as yearlings. The fate of zero-age coho is unknown (MacKay 2004, in prep.). These young coho are presumed to be of natural origin, due to hatchery practices that schedule the release coho while in their second year. In 2004, beach seine efforts yielded 190 coho juveniles, 15% of which were fish with hatchery origin. Most were yearlings, but a high percentage (~ 40%) appeared to be zero-age individuals with fork lengths less than 50 mm. Trap and seine records describe a surge in later arriving hatchery coho, following releases from Nooksack River area hatcheries. Virtually all coho released from the Kendall Creek, Skookum Creek, and Lummi Bay hatcheries display hatchery marks. Unmarked coho in the estuary gradually declined in numbers after hatchery release dates.
151
Chum Salmon Chum salmon (O. keta) have evolved to limit their freshwater life history by migrating immediately to marine waters upon hatching. As one of two of the Pacific salmon species that often spawns near river outlets, chum salmon fry do not usually require a lengthy outmigration to the sea. This life history strategy, which chum salmon share with pink salmon, reduces the mortality associated with the variable freshwater environment, but makes chum more dependent on estuarine and marine habitats (WDFW 2005). While migrating, chum fry are attracted by shade or the darkness of aquatic vegetation communities. When the density of fish becomes high in the shaded areas, they continue to move downstream (Salo 1998). This pseudo-schooling continues until they reach brackish water in the estuary. When they finally reach sea water, they respond strongly to the mixed water and either turn back to fresh water or swim in the upper layer of lower salinity (Salo 1998).
Chum salmon begin actively feeding immediately after emergence from their spawning beds, preparing for a comparatively early outmigration. Chum fry both migrate and feed at night, consequently, they predominately prey on items available this time of day (Salo 1998). Their basic diet consists of chironomid, mayfly, stonefly, and dragonfly larvae, chironomids found to be the most abundant of these benthic invertebrates (Salo 1998). Chum salmon juveniles in the estuary are small; therefore, they require small-sized prey items. Insects in larval stages comprise most of their fresh and brackish habitat diet. Between nearshore and freshwater tidal habitats, young chum feed mainly on insects, copepods, crab larvae, and the small, jelly-like invertebrate Oikopleura (Hart 1980) while foraging close to the shoreline.
Upon arrival in the estuary, chum salmon fry inhabit nearshore areas. Chum fry arriving in estuaries are initially widely dispersed, but form loose aggregations oriented to the shoreline within a few days. These aggregations occur in daylight hours only, and tend to break-up after dark, regrouping nearshore at dawn the following morning (WDFW 2005). Once in the estuary, chum fry remain a relatively short time.
Chum fry in the estuary make less use of the habitat as a nursery than their chinook counterparts; however, they have been observed to reside here for a month or more. Water temperature requirements of chum juveniles in the estuary are similar to those of Chinook [12-14°C (Brett 1952)]; however, chum fry are capable of regulating full strength seawater soon after emergence from redds, and easily assimilate the salinity gradient upon arrival to delta habitats (McNeil 2001, Salo 1998).
Aitkin (1998) notes that chum salmon are second only to chinook in dependence upon estuaries as rearing areas. Feeding in the estuary is of primary importance to chum. They are small upon entry to the estuary and must grow to a size that affords them predator avoidance in the nearshore. Chum salmon obtain their critical early growth by feeding in tidal sloughs and creeks and other intertidal areas (WDFW 2005). MacKay (2000) found that the average size (fork length) upon arrival in early April was 38.6mm. Chum fry arriving at the mouth of the river around the end of June, near the end of their fresh water migratory period, were significantly larger at 49.2mm. It is interesting to
152
note that the size of chum fry in the estuary is significantly smaller than their salmonid counterparts, with the exception of pink salmon fry. This size disparity is often a detriment to the chum fry, as they are a targeted prey item of larger Pacific salmon species co-existing in the estuary. Hence, the chum salmon are utilized as a significant food source for other fish species that feed here.
Nooksack River chum salmon hatch in the early spring and proceed immediately to the sea, arriving at the mouth of the river as early as February (MacKay 2000). Nooksack estuary sampling in 2004 revealed chum abundance that climbed rapidly in the first week in March, peaking in the first week of April, and declined by the end of May (MacKay 2004, in prep.).
Figure 81 is a plot of bi-weekly catch per seine set in 2004. There is a single peak of abundance that occurs in the first week in April. Chum juveniles were present in our seine catches in early March and were present until the first week in June.
Bi-weekly Chum Catch/Set, 2004
50.0
40.0
30.0
20.0
Average Catch/Set 10.0
0.0 1/17 1/31 2/14 2/28 3/13 3/27 4/10 4/24 5/8 5/22 6/5 6/19 7/3 7/17 Bi-Weekly Period
Figure 81. Bi-weekly beach seine catch per set for chum in 2004.
Pink Salmon Nooksack odd-year pink salmon (O. gorbuscha) are considered a unique genetic diversity unit (GDU) because they exhibit earlier river entry timing and spawn activity than other Puget Sound pink salmon stocks (Shaklee et al. 1995). This early entry and spawn time triggers an early outmigration, witnessed consistently in even-numbered years, by LNR staff at the river screwtrap. After emergence, pink salmon fry migrate quickly downstream at the stream’s surface. They spend less time, on average, in fresh water than other Oncorhynchus species. Migration duration has been documented between 53 to 72 days, depending on stream length (Heard 1998).
153
The rapid exit from natal stream habitat into estuary and marine habitats, resulting in a smaller fry fork length, turns pink fry into much sought after prey for other piscivorous species here. A principal predator on pink fry in estuaries is the Pacific herring, Clupea pallasi. Herring tend to move up into the mouths of rivers to specifically feed on migrant pink fry (Heard 1998). In addition, LNR staff observed pink fry in the mouths and stomachs of young chinook in the Nooksack estuary (Pfundt 2004, pers. comm.). So, like their chum fry counterparts, pink fry presence in the estuary is integral to the perpetuation of the resident salmonids and piscivores.
The small size of pink salmon fry upon entry to the estuary may be a factor in the exhibition of schooling behavior. Heard (1998) witnessed large schools of pink salmon fry in estuarine and nearshore habitats at night; he concluded that this schooling behavior proved an excellent defense mechanism against nocturnal predators in tidal habitats.
Due to their rapid exit from streams at emergence, pink salmon fry feed less in fresh water than other Pacific salmon (Heard 1998). Bailey et al. (1975) found chironomid pupae and other insects, as well as some plant debris in the stomachs of fry examined while still in redds, before emergence. Various research (cited in Heard 1998) found that shorter coastal streams bear pink fry that do not feed at all, whereas migration that takes several days increases prey abundance in gut contents. Larval and pupal stages of dipteran insects, particularly chironomids, are the principal food items eaten in fresh water by pink salmon fry.
After leaving fresh water, young pink salmon tend to remain close inshore through their first summer, moving into deeper water in September. At that time, they opportunistically feed on Oikopleura, amphipods, euphausiids, and young herring, eulachon (Osmeridae spp.), Pacific hake (Merluccius productus), sticklebacks (Gasterosteidae spp.), and gobies (Gobiidae spp.) (Hart 1980). Like their juvenile chum salmon counterparts, insects comprise additional prey items in the estuarine diet of pink salmon. Other items commonly found in the stomachs of pink fry migrants include larval mayflies, stoneflies, terrestrial insects, mites, and copepods (Heard 1998).
The mean size of migrant pink fry varies from 28mm to 35mm in fork length (Heard 1998). Average fork length upon arrival is about 35mm, with a minimum and maximum fork length of 26mm and 39mm, respectively (MacKay 2000). Fork lengths of 2004 pink fry in the Nooksack estuary ranged from 30mm early in the season to a maximum of 73mm by the end of May (MacKay 2004).
Between February and May of 2004, pink fry arrived at the river screwtrap near the mouth of the river. During this same time period, pink fry were captured in nearshore habitats
Estuary sampling efforts by LNR staff in 2004 revealed pink salmon fry residing in nearshore habitats from early March until the end of May (MacKay 2004).
154
Pink salmon fry arrivals to the estuary were observed between February and May in 2004 (Figure 82). They were being simultaneously caught in nearshore sampling sites, suggesting a short estuarine residence time. Whether the observed short residence in estuarine habitat is biological or environmental is unknown at this time.
Bi-weekly Pink Catch/Set, 2004
5.0
4.0
3.0
2.0
Average Catch/Set 1.0
0.0 1/17 1/31 2/14 2/28 3/13 3/27 4/10 4/24 5/8 5/22 6/5 6/19 7/3 7/17 Bi-Weekly Period
Figure 82. Bi-weekly beach seine catch per set for pink salmon in 2004.
Sockeye Salmon Due to the early return timing, low abundance, and low visibility of sockeye salmon (O. nerka), it is difficult to estimate escapement into the Nooksack River system. Sockeye salmon have long been regarded as the most commercially valuable of Pacific salmon in Canadian waters; however, regularly low abundance in the Nooksack system has not afforded this species an economically critical standing.
The typical life cycle of the sockeye salmon includes a stage of juvenile lacustrine rearing after migration from riverbed redds. However, the Nooksack River sockeye stock is a purely riverine stock; one that lacks a lake nursery in its cycle. The Nooksack River sockeye, along with its Skagit River counterpart, is not considered to be a formal stock. Recent genetic analysis of adult spawners indicates they are more closely related to river- type populations in British Columbia and Alaska than to lake-rearing populations nearby (Gustafson and Winans 1999).
Young sockeye reach the Nooksack River estuary smolt trap at age-0 and age-1, though the numbers for both age classes are very low. There were no sockeye juveniles collected during the 2003 outmigration season, and only two individuals were collected in 2004, both in late March.
Upon reaching salt water, young sockeye salmon are usually between 60 and 95mm in fork-length, but records show sockeye smolts in large Canadian rivers measuring up to
155
130mm (Hart 1980; Burgner 1998). For the early part of the summer they appear to remain inshore, within the influence of the home river (Hart 1980). While here, they feed heavily in the estuary, focusing on prey found in the nearshore and brackish environments, rarely straying back up into freshwater tidal habitats.
Once in the estuary, the larger body size of sockeye salmon affords them the opportunity to feed on a variety of prey items. Food at this stage includes crustaceans such as copepods, amphipods, decapods, barnacle larvae, ostracods, and euphausiids; insects; larval and juvenile fishes such as sand lance, rockfish (Sebastes sp.), eulachon, starry flounder (Platichthys stellatus), herring, stickleback, Pacific hake; and the larvacean Oikopleura (Hart 1980).
Sockeye juveniles are known to be heavily preyed upon by bull trout (Salvelinus confluentus), squawfish (Ptychocheilus oregonensis), rainbow trout (O. mykiss), coho salmon, and sculpin (Cottidae sp.) in estuarine and nearshore habitats (Hart 1980). Seals and gulls are notable predators of sockeye in the estuary, as well.
Anadromous Trout Steelhead trout (O. mykiss), the anadromous form of rainbow trout, inhabit all three forks of the Nooksack River. They spawn in mainstem, side channel, and tributary habitats, and produce fry that rear in freshwater for up to four years. Steelhead are unique in that they are not semelparous (commence death immediately after spawning), and as adults spending one to three years in salt water, they often return to spawn in their natal stream for a second or third time (Hart, 1980).
Nooksack River coastal cutthroat trout (O. clarkii) are of native, mixed-stock origin (Blakley et al. 2000). Though all types of cutthroat life history strategies take place in the river, only anadromous individuals spend time in the estuary. They may go to sea when quite small and take up estuarine residence for one or more years (Hart 1980). Voracious predators of salmonid fry and juveniles throughout the river, young coastal cutthroat serve as an important food source to those same prey species. This predator-prey interaction is always size-specific; the larger fish will always prey on the smaller fish.
Steelhead and coastal cutthroat trout spend variable time in the estuary, consequently, their diet in the estuary is diverse. It is shaped by the size and energy requirements of individual fish. These anadromous trout may be one to four years old upon first entry into the estuary. Considering that the estuary boasts a large, diverse food web, food may not be limiting to these fish.
Once in the estuary, trout may stay here for up to a year, feeding heavily on other fishes such as coho salmon, stickleback, rockfish, sculpin, and flatfishes. Smaller individuals regularly eat crustaceans, and both aquatic and terrestrial insects (Hart 1980).
The Nooksack River is thought to support two stocks of steelhead, a summer-run stock, and a winter-run stock. Both stocks are native, but have unknown stock status (WDFW & WWTT 1993). Winter-run adults usually return to their natal streams between
156
November and May, spawning from January to June, and summer-run steelhead escapement lasts from May to October. Summer-run stocks spawn between February and April (Anchor Environmental 2001). No official escapement is estimated for these fish in the Nooksack; however, Nooksack Natural Resources (Currence, pers. comm., 2004) puts escapement of steelhead into the Nooksack system between 100 and 400 adults.
Due to the unconventional spawn timing and migration patterns of steelhead in the Nooksack basin, the sizes of individuals, either young smolts or adult kelts arriving at the lower river trap, range from 70mm to 700mm (MacKay 2000). It is also difficult to distinguish winter from summer-run steelhead juveniles, as both stocks tend to leave the river year-round. LNR catch records at the rotary screw trap in the estuary between 1994 and 1999 indicate that steelhead juveniles outmigrate through the lower river during all months of operation, with a peak between May and June (MacKay 2000). Dispersal patterns of trout once in the estuary are unknown at this time.
Native char The U.S. Fish and Wildlife Service issued a final rule listing the Coastal-Puget Sound bull trout (Salvelinus confluentus), a distinct char population segment, as threatened on November 1, 1999 (64 FR 58910). The Nooksack drainage is one of four in Puget Sound that supports a viable, wild population of anadromous bull trout. Although considered a threatened species by the U.S. Fish and Wildlife Service, this fish is found throughout the Nooksack basin, in all three forks, and the mainstem down to the estuary.
Also unique to this population segment is the overlap in distribution with Dolly Varden (S. malma), another native char species extremely similar in appearance to bull trout, but distinct genetically (N. Currence, pers. comm.). Once thought to be a single species, the two are formally recognized as separate. One important factor distinguishing the two from each other should be noted. Bull trout are migrants, much larger in size, piscivorous, and appear to dominate the mainstems of natal rivers. Current evidence suggests that the Dolly Varden in Washington are distributed as isolated tributary populations above natural anadromous barriers, while bull trout tend to be distributed below these barriers and are often anadromous (WDFW 1998; Spruell and Maxwell 2002). Based on this information, all native char observed in accessible anadromous reaches are believed to be bull trout.
Native Nooksack char are among the most aggressive predators of young salmon; therefore, they play a significant role in shaping juvenile salmonid populations in this system. Their population is not profuse, but their numbers have remained consistent in the drainage over the past twenty years (Dunphy, pers. comm.). Seaward migration takes place in the spring, after a three-year maturation period in fresh water (Hart 1980). The fish are usually between 170 – 190 mm in fork length upon entry to the estuary. Dolly Varden, however, do not leave the estuary, rather, they spend a short time here and head back upstream to spawn in the fall.
157
Common prey items of char in the estuary include small herring, stickleback and young salmon; salmon eggs, mollusks, insects, and Crustacea (Hart 1980). Additionally, sand lance, surf smelt, and shiner perch provide food for bull trout in nearshore environments.
The occurrence of Dolly Varden or char in the Nooksack estuary trap and seine catches was rare. In 2004, one individual, presumed to be bull trout, was caught in the first week in May. Peak abundance in the estuary was mid to late June, declining in July (Figure 83).
Bi-weekly Char Catch at Primary Sites, 2004 15
10
5 Average Catch/Set
0 1/17 1/31 2/14 2/28 3/13 3/27 4/10 4/24 5/8 5/22 6/5 6/19 7/3 7/17 Bi-Weekly Period
Figure 83. Bi-weekly beach seine catch per set for char in 2004.
Non – Salmonids Baitfish, or those species that are known prey items for salmonids, catch totals from 2003 and 2004 are described in Table 9. Catch effort differed between the two years sampled: 1,065 seine sets were made in 2003; 864 sets were made in 2004. Longfin smelt were only found in 2004, and were much less common than herring, surf smelt, and sandlance that were sampled.
Table 9. Nooksack estuary bait fish species sampled by beach seine in 2003 and 2004. Species 2003 2004 Herring 212 215 Surf Smelt 566 159 Sandlance 606 142 Longfin Smelt 0 15
158
Pacific Herring Most herring sampled were post-larval forms, ranging in length between 78 - 91mm. In 2003, nearly all herring sampled were taken from an exposed nearshore site off of Portage Island in May; near the Nooksack Delta in early June, and in the Squalicum Creek estuary in late June. One of the largest catches of herring in beach seining efforts landed 200 individuals near Portage Island in April of 2004 during a 6.0-foot tide.
Surf smelt In 2003, surf smelt catches were most abundant in early March and late June, ranging from Gooseberry Point nearshore and Brant Spit off of Portage Island, respectively. Several dozen were sampled between April and late May at the mouths of Squalicum and Padden Creeks. In 2004, surf smelt catches were highest in late March, concentrated at the mouth of the Lummi River. Most individuals sampled were adults, but several post- larval forms were also caught (lengths 52-85mm).
Sandlance Nearly all sandlance sampled with beach seine gear were post-larval forms, with lengths ranging between 35 and 86mm. In 2003, sandlance were caught in beach seine exercises between late March and late June. Individuals caught earlier in the season were found at the mouth of the Lummi River and along the Gooseberry Point nearshore; in June, concentrations were found along Bellingham Bay nearshore sites, and at the mouth of Squalicum Creek. Toward the end of June, large populations were sampled in the exposed shoreline habitat between Cherry Point and Sandy Point. The highest catch in a single set, 44 individuals, was sampled near the Padden Creek estuary in July of 2004. Nearly 80% of that year’s samples were caught in late July.
Longfin smelt Longfin smelt were captured between March and April in 2004, at sites dominated by fresh water. Delta channel habitats were the most common places that longfin smelt were caught; however, several were sampled in the mainstem channel just below Marine Drive Bridge, and from an upstream site near the Slater Road Bridge.
159